A novel layered topology of auxetic materials based on the

tetrachiral honeycomb microstructure

Ferdinando Auricchio1, Andrea Bacigalupo2*, Luigi Gambarotta3, Marco Lepidi3,
Simone Morganti1, Francesca Vadalà3
1
Department of Civil Engineering and Architecture, University of Pavia, Italy
2
IMT School for Advanced Studies Lucca, Italy
3
Department of Civil, Chemical and Environmental Engineering, University of Genova, Italy

Abstract
Microstructured honeycomb materials may exhibit exotic, extreme and tailorable mechanical
properties, suited for innovative technological applications in a variety of modern engineering
fields. The paper is focused on analysing the directional auxeticity of tetrachiral materials,
through analytical, numerical and experimental methods. Theoretical predictions about the
global elastic properties have been successfully validated by performing tensile laboratory tests
on tetrachiral samples, realized with high precision 3D printing technologies. Inspired by the
kinematic behaviour of the tetrachiral material, a newly-design bi-layered topology, referred to
as bi-tetrachiral material, has been theoretically conceived and mechanically modelled. The
novel topology virtuously exploits the mutual collaboration between two tetrachiral layers with
opposite chiralities. The bi-tetrachiral material has been verified to outperform the tetrachiral
material in terms of global Young modulus and, as major achievement, to exhibit a remarkable
auxetic behaviour. Specifically, experimental results, confirmed by parametric analytical and
computational analyses, have highlighted the effective possibility to attain strongly negative
Poisson ratios, identified as a peculiar global elastic property of the novel bi-layered topology.

Keywords: Elastic properties, Experimental tests, Additive manufacturing, Finite element

analysis, Beam lattice, Bi-tetrachiral material

____________________________________________
*
Corresponding Author
INTRODUCTION

Architectured composite materials characterized by a cellular honeycomb microstructure are

experiencing an increasing success in a variety of innovative technological applications, by
virtue of superior and tailorable physical-mechanical properties [1],[2],[3]. Among the other
cellular topologies, honeycomb materials based on chiral and anti-chiral microstructure can
offer peculiar extreme features of directional auxeticity, frequently conjugated with functional
static performances of high resistance to fracture and indentation, atypical bending with
synclastic curvatures, enhanced strength against buckling instabilities [4],[5],[6],[7],[8],[9].
Furthermore, the fine parametric tunability of the microstructural inertia and stiffness allows
the systematic employment of chiral and anti-chiral materials to realize efficient and versatile
phononic filters, elastic waveguides and acoustic diodes [10],[11],[12],[13],[14]. Within this
challenging framework, the optimal design of the micromechanical properties opens interesting
research perspectives towards the theoretical conceptualization and experimental validation of
a new generation of smart materials targeted at innovative engineering applications, including
impact absorption, negative refraction, shape morphing, wave trapping, vibration shielding,
noise silencing and invisibility cloaking [15],[16],[17],[18],[19],[20].
Focusing the attention on the modern scientific literature about quasi-static laboratory tests
of architectured materials, a pioneering investigation has experimentally verified the theoretical
possibility to achieve strong auxetic properties (quantified by negative Poisson ratios close to
unity) in medium-size samples of planar hexachiral honeycombs hand-built in polystyrene [21],
[22]. Chiral and anti-chiral topologies of planar honeycombs have been analyzed by means of
quasi-static experimental tests on small-size samples manufactured using selective laser
sintering rapid prototyping of nylon powder. The experimental results have been successfully
compared with the static response simulated by finite element models and simplified analytical
formulations [23]. Satisfying agreement has been achieved in terms of global elastic properties
(Young modulus and Poisson ratio) for almost all the studied topologies, including those
exhibiting auxeticity along certain tested directions. The research findings have left some open
issues regarding the tetrachiral material, exhibiting the highest discrepancy between numerical
results and experimental evidences. In the same research field, successful comparisons between
experiments and simulations on medium size samples have been also obtained for (i) rapidly
prototyped chiral honeycombs in terms of transverse shear stiffness [24], (ii) laser-crafted re-
entrant anti-trichiral honeycombs in terms of Young moduli and Poisson ratios [25], (iii)
selective laser sintered tetrachiral and hexachiral honeycombs in terms of flatwise buckling
loads [26]. The promising outcomes of experimental investigations on smart tetrachiral and
hexachiral honeycomb have demonstrated the viability of advanced applications in the
structural health monitoring and other sensing purposes [27].

2
 More recently, honeycomb sheets with chiral and anti-chiral cellular topologies have been
employed to develop three-dimensional curved structures. The auxeticity has been numerically
simulated in microstructured cylindrical stents and experimentally verified in planar steel
samples fabricated through the waterjet cutting technology [28]. Finite element results and
experimental measures have been also compared to assess the coupled extensional and torsional
deformations of microstructured cylindrical shells fabricated using selective laser sintering and
selective laser melting methods with nylon and aluminium alloy [29]. Within this comparative
study, the effects of a reinforcing rotation disk introduced at the cylinder mid-height have been
also considered [30]. High-pressure abrasive waterjet technologies have been applied to realize
functionally graded tetrachiral structures. The influence of the geometric gradient factor on the
structural elastoplastic response has been numerically predicted and experimentally analysed
during compressive load tests [31].
Negative Poisson ratios have been experimentally observed in hybrid chiral bi-dimensional
materials fabricated via multi-material 3D printing. The auxetic behaviour, combined with
sequential cell-opening mechanisms, has been purposely designed to develop innovative multi-
functional composites, characterized by smart sensitivity to environmental conditions and
targeted to technological applications in drug delivery and colour changing for camouflage [32].
New topologies thought for 3D auxetic material have been based on multi-layered tetrachiral
schemes, with inter-layer clockwise and anticlockwise ligament connections. By tuning the
direction of the interlayer connections, materials with two positive and one negative Poisson
ratios have been first analytically predicted and successively confirmed through numerical
simulations and experimental test on medium-size samples printed with the stereolithography
technology [33]. Other different topologies of 3D chiral materials have been proposed and their
deformation mechanisms have been experimentally studied through tensile and compression
tests on selected laser sintered samples [34].
The most notable trend emerging from the review of the most recent state-of-the-art is the
increasing and pervasive employment of additive manufacturing solutions for the engineering-
oriented application of the large amount of theoretical knowledge about auxetic materials based
on tailorable microstructures [35],[36],[37],[38]. Indeed, additive manufacturing is rapidly
evolving as one of the most promising manufacturing technologies for designing, optimizing,
rapid prototyping and large scale producing three-dimensional architected cellular materials
with high-fidelity realization of complex microstructural topologies [39], [40], [41], [42], [43],
[44]. Interesting advanced applications for additive manufactured architected materials range
across many modern fields in frontier engineering, from micro-electro-mechanical systems to
lightweight components for automotive or aerospace industry, from patient-specific medical
implants to smart structural elements in parametric engineering and architecture.

3
 The additive manufacturing process always starts from a virtual 3D model that has to be
converted into a 3D printing-suitable format (the most common is the Standard Triangulation
Language). Then, a so-called slicing procedure is performed and, for each slice, specific
machine instructions are defined, which govern the 3D printer during the layer by layer
production process. Among the others, one of the most widespread, versatile, and economic 3D
printing processes is the Fused Deposition Modelling (FDM). This technology employs a
thermoplastic material that is first heated to a semi-molten state and then extruded through a
robotically-controlled nozzle in a temperature-controlled environment to construct layer by
layer the desired part. Currently, a large variety of thermoplastic materials can be extruded,
ranging from acrylonitrile butadiene styrene (ABS) and polylactic acid (PLA) to techno-
polymers, like polyetherether ketone (Peek) or polytherimide (Ultem). Thanks to their superior
mechanical properties, these thermoplastics can be used to produce also structurally functional
components that, by virtue of the relatively low-cost of the technology, have been used for a
wide spectrum of innovative technological applications ranging from acoustics and mechanics
[45],[46],[47] to biomedicine and pharmaceutics [48],[49], from electronics [50],[52] to social
applications [53],[54].
This stimulating and challenging scenario motivates the leading idea to conjugate the most
recent progresses in additive manufacturing with the pressing demand to establish a robust
experimental background supporting the most advanced theoretical and applied researches in
the exotic elasticity and smart engineering functionality of existing and new architectured
materials. According to these basic motivations, the paper leverages the actual technological
possibility in realizing high-fidelity complex topologies by 3D printing thermoplastic materials
in order to bridge the scientific gap between analytical or numerical predictions and
experimental evidences in the field of chiral microstructured materials. Focus is laid on
experimentally validating some theoretical results about the directional auxeticity of the
tetrachiral material [25],[55],[56]. In this respect, planar polymeric samples have been 3D
printed by employing the FMD technology (Section 1), in order to realize a tetrachiral cellular
geometry with the highest possible precision (Section 1.1). Following a multidisciplinary
approach for the data acquisition and processing, the samples have been tensile tested and the
quasi-static response has been measured by means of non-contact technologies based on digital
image acquisition (Section 1.2). Therefore, the measures have been analysed by virtue of
numerical data post-processing to solve the inverse problem concerned with the input-output
identification of the global elastic properties (Section 1.3). In parallel, different mechanical
models of the tetrachiral samples have been developed (Section 2), in the framework of solid
mechanics (Section 2.1) and structural mechanics (Section 2.2). Analytical and numerical
solutions have been determined to simulate the experimental tests and compare the respective
findings in terms of global Young modulus and Poisson ratio (Section 2.3). As valuable point

4
of novelty, the research outcomes have led to the theoretical conceptualization, mechanical
modelization and analytical/numerical simulation of an original by-layered topology, based on
tetrachiral layers. The new topology, which differs from other layered auxetic materials based
on radially foldable microstructure [57] and does not require a different bi-layered multi-
material 3D printing process, is kinematically based on the independent and opposite-sign
rolling up mechanisms of the component layers, reciprocally constrained at the boundaries. The
theoretical predictions and the experimental behaviour have been compared in terms of global
rigidity and auxeticity (Section 3). Finally, concluding remarks have been pointed out and
future developments have been outlined.

1. TETRACHIRAL SAMPLE

The class of chiral and antichiral cellular materials is characterized by a periodic tessellation of
the bidimensional plane. The elementary cell is strongly characterized by a microstructure
composed by stiff circular rings connected by flexible straight ligaments, arranged according
to different planar geometries including the trichiral, hexachiral, tetrachiral, anti-trichiral, anti-
tetrachiral topologies. Among the others, the tetrachiral material is featured by a monoatomic
centrosymmetric cell in which the central stiff and massive ring (or disk) is connected to four
tangent flexible and light ligaments.

1.1 Sample preparation

A polymeric sample of the tetrachiral geometry has been realized with Fused Deposition
Modelling (FDM) technology by the Group of Computational Mechanics and Advanced
Materials of the University of Pavia. For the layer-by-layer FDM preparation of the samples a
thermoplastic filament made of the polymer Acrylonitrile Butadiene Styrene (ABS) has been
used. This material is known to have a Young modulus ranging in the large interval 1100-2900
MPa [58]. In the initial pre-print conditions, the nominal value given by the filament
manufacturer is close to 2000 MPa. The reduced value for the post-print conditions is a matter
of experimental identification and has been tentatively fixed at 1300 MPa as initial realistic
value. The Poisson ratio has been realistically fixed at the nominal value of 0.35.
The printing head movements and all the printing parameters are automatically controlled
by an electronic board relying on a set of instructions (i.e., the G-Code). The G-Code is
produced by a dedicated software, commonly called slicer or slicing software that takes into
account the virtual geometry, the properties of the printing material, and the specific features
of the 3D-printer. The 3D printer used to produce the tetrachiral sample was a 3NTR A4v3 (see
Figure 1a). The machine has been equipped with three extruders that can be heated up to 410
°C; a nozzle of 0.4 mm of diameter has been used. The build-tray temperature has been set to
110 °C, while the heated chamber temperature to 70 °C in order to avoid distortions of the
printed sample induced by the high thermal gradients occurring during the manufacturing
5
Figure 1. Tetrachiral sample printed with the FDM technology: (a), (b) 3NTR 3D printer preparing the
tetrachiral sample; (c) 3D printed sample; (d) geometry of the printed sample.

process. A filament cross-section with thickness of 0.2 mm and width of 0.4 mm, and a fiber-
to-fiber overlap of 0.04 mm has been assumed. In Figure 1b the printed sample is shown. The
geometric dimensions of the printed sample are reported in Figure 1c and detailed in Table 2.

Table 1. Geometric properties of the tetrachiral and bi-tetrachiral samples

Dimensions (in mm) Notes

Inter-ring distance H 20 -
Ring mean radius R 4 -
Ring width w 0.8 -
Ligament width w 0.8 -
Sample depth d 20 Tetrachiral sample
Sample depth d 6 Layer of the bi-tetrachiral sample

6
(a) (b)
Supporting truss system

First
Row

4‐by‐6 
5H inner
cluster

Last 
Row

Equal
forces
5H

Figure 2. Experimental set up in the initial reference configuration: (a) picture, (b) sketch of the
supported tetrachiral sample under the action of the external forces.

1.2 Experimental tests

The experimental activities have been carried out at the Laboratory of Structural Engineering
of the DICCA – University of Genova. The tetrachiral samples made of an array of 6-by-6 cells
have been tested under uni-axial tension according to a force-control scheme, within stable
environmental conditions (Figure 2a). The test set up has been designed to apply known
increments of force to the sample, starting from an initial reference configuration.
In the reference configuration, the sample is hanging vertically under self-weight load and
supported by a constraining system applied at the top side (Figure 2b). The constraining system
has been purposely designed to ideally preserve the alignment of all the rings located in the first
(top) row of the sample (yellow-filled black rings in Figure 2b). The transversal displacements
and rotations of all the rings remain ideally free, allowing the unrestricted development of both
lateral expansion/contraction and the rolling-up mechanism. The controlled force acts at the
bottom side of the sample by means of a steel truss system conveniently designed to split the
total vertical action into equal forces (red vectors in Figure 2b) applied at all the rings located
in the last (bottom) row of the sample (pink-filled black rings in Figure 2b). In order to assess
the global elastic properties of the tetrachiral sample, the 4-by-6 inner cluster of internal
unconstrained and unloaded cells (green region in Figure 2b) been considered in the following.
The uni-axial tension test has been run by applying five increments of the quasi-static force
F2 , corresponding to equivalent global stress  22  F2 A2 , where A2  6 Hd is the cross-
section area of the ideal solid (rectangular parallelepiped) with dimensions 6H (width), 6H
7
(height) and d (depth). The force increments (steps 1-5) are reported in Table 2, starting from
the initial loading conditions (step 0), under the self-weight of the sample and the truss system.
Particular attention has been payed to some operational issues, like preserving the vertical
planarity of the deformed configurations and minimizing the parasitic effects of friction in the
constraints. Finally, the entire loading process has been designed not to overcome the limit of
linear reversible deformation in the material of the sample. This design requirement has been
checked a posteriori by verifying that the initial undeformed configuration is recovered at the
end of the unloading process, here not reported for the sake of synthesis.

Table 2. Loading steps for the uni-axial tension test of the tetrachiral sample

Force F2 [N] Global stress  2 2 [N/m2]

Step 0 8.15 4075

Step 1 18.15 9075
Step 2 28.15 14075
Step 3 38.15 19075
Step 4 48.15 24075
Step 5 58.15 29075

1.3 Data processing and identification

1.3.1 Image acquisition and post-processing
Among the different available possibilities for data recording in quasi-static experimental
tests, the non-contact data acquisition through a digital camera has been selected as a convenient
compromise between measure reliability and operational feasibility [59]. Specifically, one or
more two-dimensional images (with dimensions 5184-by-3456 pixels) of the tetrachiral
specimen have been acquired for each load step with a fixed camera (14 bit Canon EOS 600D,
with image processor DIGIC 4). In order to minimize distortion, the focal plane of the digital
camera has been initially calibrated to be parallel to the specimen plane, with the focal axis
crossing the geometric center of the specimen in the undeformed configuration.
All the digital photographs have been uploaded and converted in two-dimensional arrays
of pixel coordinates within the Matlab environment [60], in order to be post-processed by using
the Image Processing Toolbox. First, the background grid of known dimensions (approximately
co-planar with the specimen) has been employed to convert the pixel coordinates of the digital
photograph into a real coordinate system. Second, the real positions x of the ring centers have
been assessed by programming an automatic function for the recognition of assigned shapes,
based on the Hough transform. This mathematical transform is an efficient tool for geometric
shape recognition, largely used in digital image processing and computer vision. The automatic
8
Figure 3. Non-contact image-based identification of the real coordinates for all the ring centers in the
inner cluster of the tetrachiral sample: (a),(b) loading step 0; (c),(d) loading step 5.

recognition of an assigned shape is based on searching object imperfections within a certain

class of geometric objects, in which the most suited candidates are recognized as local maxima
in a finite parameter space (also known as accumulator space). Specifically, the Hough
transform is classically concerned with the identification of lines in digital images, but it can
be easily extended to identify positions of arbitrary curves. In particular, the recognition of
circumferences is commonly used to identify the centers and radii of circles [61].
By properly tuning the sensitivity parameters of the automatic function, the identification
of the real coordinates of all the unconstrained ring centers in the tetrachiral specimen has been
successfully run for each load step (Figure 3). Therefore, the total displacements of the ring

9
 (a) (b)
H H H H H 5HE11

Inner cluster

H Reference
configuration

3HE22

Deformed configuration

Figure 4. Identification of the global elastic properties: (a) affine displacement field and physical
meaning of the normal strains 11 ,  22 related to the deformed configuration; (b) comparison between the
experimental deformed configuration at the loading step 5 (black crosses) and the corresponding
identified deformed configuration (red circles and red displacement vectors)

centers have been calculated as position differences with respect to the initial load step, and the
results have been stored in the experimental set of punctual displacement vector v .

1.3.2 Input-output identification of the global elastic properties

According to a mechanical model-based identification approach, the global elastic properties
of the tetrachiral samples are identified by introducing the in-plane affine displacement field u
of an equivalent homogeneous Cauchy continuum in the framework of a linear kinematics
(Figure 4a). Accordingly, the continuum displacement field u(x) of the material point at the
generic position x is expressed through the classic relation
u(x)  u0  Hx (1)

where u0 is the displacement of an arbitrary origin point. The two-by-two H -matrix is the
displacement gradient, composed by a skew-symmetric part W and a symmetric part E that
can be recognized as infinitesimal rotation matrix and infinitesimal strain matrix, respectively.
The experimental set of punctual displacement vector v measured at known positions x
can conveniently be supposed – as first approximation – to obey the linear vector law
v(x)  v0  Gx (2)

where v0 is the known displacement of a measurement point taken as origin and the two-by-
two coefficient matrix G of linear proportionality is unknown a priori. Since the displacement
variable v is experimentally known only in a finite number m of measurement points (Figure
4b), the identification issue consists in imposing m relations

10
 vh  v0  Gxh , h  1,..., m (3)

and searching for the four independent G -components. Once the G -components have been
determined, the symmetric and skew-symmetric parts of the G -matrix can be extracted.
Finally, by recognizing the formal analogy between the equations (1) and (2), the two
infinitesimal rotation and infinitesimal strain matrices can be univocally identified
E  H  H   G  G 
1
2
T 1
2
T

(4)
W   H  H   G  G 
1
2
T 1
2
T

Therefore, once the matrices E and W are identified, the normal strains E11 , E22 , the angular
strain E12 and the infinitesimal rotation W21  W12 read
 11  G11 ,  22  G22 ,  12  1
2  G12  G21  , W12  1
2  G12  G21  (5)

Furthermore, the global elastic properties can be assessed by the well-known relations
 22 H11
E ,   (6)
H 22 H 22

where E and  are the global Young modulus and global Poisson ratio, respectively.
From the mathematical viewpoint, the identification problem can be properly formulated
by stating the algebraic linear equation
Ay  b (7)

where y   G11 , G22 , G12 , G21  is the four-by-one vector of unknowns, A is a known 2m -by-
four matrix depending only on the positions of the measurement points and b is a known 2m -
by-one depending only on the measured displacements.
In the common operational case corresponding to redundancy of measures ( m  3 ), the
algebraic problem turns out to be overdetermined. Therefore, it is necessary to solve equation
(7) according to the least square approximation of the solution, which reads
y  A+ b (8)

where A  stands for the pseudoinverse of the rectangular matrix A .

2. NUMERICAL SIMULATIONS AND ANALYTICAL PREDICTIONS

2.1 Solid finite element model
The tetrachiral samples can be modeled according to a high-dimensional formulation in the
framework of linear solid mechanics. The software Autodesk Inventor [62] has been used as
CAD modeling tool to describe with high-fidelity the complex three-dimensional geometry of
the generic tetrachiral samples composed by a bidimensional array of N  M cells, with N
rows and M columns (Figure 5a illustrates the six-by-six cell sample). Therefore, the software
11
 Figure 5. Three-dimensional solid model of the tetrachiral sample: (a) perspective geometric view of
the six-by-six cell sample; (b) detail of the finite element mesh for the periodic cell, (c),(d) selection of
finite element nodes for the displacement reconstruction of the ring centers in the inner cluster.

Abaqus Standard [63] has been used as finite element mesh generator and solver. The entire
sample domain has been discretized with 8-node linear brick, reduced integration, hourglass
control (C3D8R) elements. The six-by-six cell sample has been meshed in 146220 brick
elements and the mesh detail of the periodic cell are illustrated in Figure 5b.
The solid model has been specified by assuming isotropic homogeneous material with
elastic properties described by the Young modulus E and the Poisson ratio  . To reproduce
the operational conditions of the experimental tests, rigid body definitions have been applied to

12
(a) (b) (c) q5
5
First
Row
ligament
ring q4
q2

N rows

2 1
H
4
q1
2R
Last  q3
Row

3
M columns H

Figure 6. Tetrachiral sample: (a) repetitive planar pattern, (b) periodic cell, (c) beam lattice model.

the nodes belonging to the internal surface of all the rings located in the first (top) row and the
last (bottom) row of the sample. The boundary conditions at the top side of the tetrachiral
sample have been imposed by applying external constraints at the rigid bodies of the first (top)
cell row. Specifically, all the vertical displacements (plus one horizontal displacement to avoid
rigid motion solutions) are constrained. Equal vertical forces have been applied to each centroid
of all the rigid bodies of the last (bottom) cell row.
From the finite element solution, the 48-by-one displacement subvector v that collects the
horizontal and vertical displacement of the ring centers in the inner cluster of 4-by-6 cells has
been reconstructed and subsequently employed to identify the global elastic properties.
Consistently with the experimental measurements, the displacement reconstruction for all the
ring centers in the inner cluster has been achieved by properly averaging the displacements of
all the finite element nodes at the internal ring circumference (see red nodes in Figure 5c,d).

2.2 Beam lattice model

The periodic cell of the tetrachiral samples can be modelled according to a low-dimensional
lagrangian formulation in the framework of linear mechanics (Figure 6a). The tetrachiral planar
geometry is characterized by the side length H of the periodic square cell and the mean radius
R of the rings (Figure 6b). According to simple trigonometric considerations, the chirality
angle of the tangent ligaments is   arctan  2 R H  and the ligament length is L  H cos  .
The lagrangian model is synthesized on the base of a few mechanical assumptions. The
stiff rings are conveniently described as annular rigid bodies, while the flexible ligaments are
described as unshearable beams. A linear elastic material with Young modulus E is assumed
for all beams. Therefore, each beam is characterized by axial rigidity EA and in-plane flexural
rigidity EJ . Moreover, the beam-annulus connections are assumed perfectly rigid joints.

13
 The deformed configuration of the periodic cell is fully described by three planar degrees-
of-freedom for each configurational node. An internal configurational node is located at the
annulus centroid, while four external configurational nodes are located at the beam midspan,
where the cell boundary crosses the ligaments (Figure 6c). The degrees-of-freedom of the
internal node and i-th external node (with i  1,..., 4 ) are collected column-wise in the
generalized displacement vector q1   u1 , v1 , 1  and qi   ui , vi ,i  , respectively, where u
stands for the rightward horizontal displacement, v for the upward vertical displacement,  for
the counter-clockwise in-plane rotation. All the nodal degrees-of-freedom are collected in the
15-by-1 cellular displacement vector q   q1 , q 2 , q 3 , q 4 , q 5  . Following the direct stiffness
method, the 15-by-15 stiffness matrix K of the generic cell can be determined. The non-zero
submatrices of the stiffness matrix K are reported in Appendix A.
Considering the tetrachiral sample composed by a bidimensional array of N  M cells
(with N rows and M columns), the global 3P -by-one displacement vector q g can be defined,
where the total number of configurational nodes in the unconstrained lagrangian model is
P  3NM  N  M . Introducing the cellular-to-global change-of-coordinates q j  Λ j q g for the
j -th cell (with j  1,..., NM ), the 3P -by- 3P global stiffness matrix reads
NM
K g   Λ Tj KΛ j (9)
j 1

where Λ j is the 15-by- 3P boolean allocation matrix. Consequently, the algebraic equations
governing the static equilibrium of the unconstrained lagrangian model are
K gqg  fg (10)

where fg is the 3P -by-one vector of the forces acting in the components of the vector q g .
In order to describe the boundary conditions at the top side of the tetrachiral sample,
opportune external constraints of the central nodes of the first (top) cell row must be imposed.
Defining the  M  1 -by-one displacement vector qc that collects all the vertical displacements
(plus one horizontal displacement to avoid rigid motion solutions) of the central nodes in the
M cells of the first (top) row, the following partition can be introduced
K ff K fc   q f   f f 
K K       (11)
 cf cc   q c   fc 

where the  3P  M  1 -by-one vector q f collects all the free degrees-of-freedom, playing the
role of lagrangian coordinates. Indeed, imposing the constraint condition qc  0 , the equation
(11) can be inverted to obtain the static equilibrium solution
q f  K ff1f f
(12)
fc  K cf K ff1f f

14
 Figure 7. Finite element simulation of the tetrachiral sample: (a) color map of the horizontal
displacement component; (b),(c) comparison between the undeformed (gray lines) and deformed
configuration (yellow solid).

where ff has non-null equal components corresponding to the only vertical forces acting in the
central nodes of the last (bottom) cell row. In order to extract from the displacement solution
q f the pseudo-experimental results necessary for the identification of the global elastic
properties, the 48-by-one displacement subvector v that collects the horizontal and vertical
displacement of the central nodes in the inner cluster of 4-by-6 cells is considered.

2.3 Comparison of experiments and simulations

The experimental results are compared qualitatively and quantitatively with the simulations
obtained with the lagrangian model and with the finite element solid model of the tetrachiral
sample. First, from a qualitative point of view, all the experimental deformed configurations
(see for instance Figure 3c) and the corresponding static simulations agree in exhibiting a non-
symmetric behaviour in response to the application of the symmetric scheme (with respect to
the vertical in-plane axis crossing the sample barycenter) of external forces. This behaviour can
be essentially attributed to the chirality of the cellular topology, which determines the
development of local deformation mechanisms, activated by the rotation of the rings around
their centers (rolling up). Globally, the vertical stretching of the sample – collinear to the
external force direction – is accompanied with an evident and significant angular strain. This
characteristic behaviour is observable in the deformed configurations obtained from the
experimental test (Figure 3d), as well as in the numerical simulations obtained from the solid
model (Figure 7). It is worth noting that the sign of the angular strain is univocally related to
the sign of the chirality angle in the cellular topology (angle  in Figure 6b). For the numerical
15
(a) (b) (c)

Figure 8. Comparison between experimental results (red circles) and numerical results of the solid model
(blue and green circles) and beam lattice model (yellow circles): (a) equivalent global stress  2 2 versus
normal strain  2 2 , (b) normal strain 11 versus normal strain  2 2 , (c) angular strain 12 versus normal
strain  2 2

simulations, the horizontal component of the normalized displacement field U Umax is shown
in Figure 7a, while the deformed and undeformed configurations are compared in Figure 7b,c.
Second, from a quantitative point of view, the normal and angular strains identified starting
from the experimental data are compared with the corresponding strains identified starting from
the numerical results obtained with the Lagrangian and the finite element solid models,
respectively (Figure 8). As a major remark, it can be noted that the experimental response (red
circles) shows – with a good approximation – a linear behaviour under increasing values of the
external forces. With focus on the solid model, it can be observed that the ratio  22 E22
obtained from the numerical results (blue circles in Figure 8a) systematically underestimates
the corresponding experimental ratio (red circles in Figure 8a). This occurrence has
demonstrated the need to update the nominal value initially assumed for the Young modulus of
the ABS material (1300 MPa), which has been recognized as the most uncertain mechanical
parameter according to the initial information available. Therefore, this parameter has been
properly updated in order to zeroing the difference between the global Young modulus E
identified from the numerical results (load-independent value marked by green circles in Figure
9a) and the average of the global Young modules ( E  1.02 MPa), identified from the
experimental results at different loading steps (red circles in Figure 9a). The updating procedure
has required an increment in the Young modulus of the ABS material of about 18% (up to 1540
MPa). The ratio  22 E22 obtained from the numerical results of the updated solid model (green
circles in Figure 8a) shows a better agreement with the corresponding experimental ratio (red
circles in Figure 8a).
Adopting the updated solid model, a very satisfactory agreement is achieved between the
ratio E11 E22 obtained from the numerical results (green circles or, rigorously, the slope of the
dashed line connecting the green circles in Figure 8b) and the corresponding experimental ratio

16
 (a) (b)

Figure 9. Comparison between experimental results (red circles) and numerical results of the solid
model (green circles) and beam lattice model (yellow circles). Global elastic properties versus external
load F22 : (a) Young modulus E , (b) Poisson ratio  .

(red circles or, rigorously, slope of the dashed line linearly regressing the red circles in Figure
8b). The negative value systematically attained by this ratio for increasing external forces
demonstrates a non-auxetic behaviour of the tetrachiral sample in the direction orthogonal to
the external forces. This key result is effectively synthesized by the positive values
systematically identified for the global Poisson ratio  (Figure 9b). In particular, the load-
independent value (   0.27 ) identified from the numerical results (green circles in Figure 9b)
closely matches the mean value (   0.26 ) identified from the experimental results (red circles
in Figure 9b). It is worth noting that these two consistent values are also in good agreement
with the analytical results obtained from a second gradient continuum model of the tetrachiral
material formulated according to a proper homogenization technique [55]. Finally, the first
qualitative remark concerned with the development of a non-negligible angular strain is
quantitatively confirmed by the identification of the strain E12 that assumes experimental and
numerical values in mutual agreement and quantitatively comparable with the values of the
normal strain E22 (green and red circles in Figure 8c).
With focus on the Lagrangian model, a good matching is found in the simulation (yellow
circles in Figure 8a) of the experimental ratio  22 E22 . Differently, the global Young modulus
identified from the Lagrangian simulation (load-independent value marked by yellow circles in
Figure 9a) returns a slight underestimation ( E  0.94 MPa) of the mean experimental value (red
circles in Figure 9a). This difference is a well-established finding that can be attributed to the
rigid body assumption for the central ring and to the overestimation of the effective length in
the flexible ligaments composing the cellular microstructure [55]. Furthermore, the simplifying
17
(a) (b)
Supporting truss system

First
Row

4‐by‐6 
5H inner
cluster

Last 
Row

Equal
forces
5H

Figure 10. Experimental set up in the initial reference configuration: (a) picture, (b) sketch of the
supported bi-tetrachiral sample under the action of the external forces.

assumptions of the Lagrangian formulation do not allow the accurate assessment of the
experimental value for the ratio E11 E22 (compare red and yellow circles in Figure 8b). Indeed,
it can be highlighted that the corresponding load-independent identification of the global
Poisson ratio (   0 marked by yellow circles in Figure 9b) is perfectly consistent with the null
value (   0 ) exactly obtainable by a micropolar continuum model of the tetrachiral material
formulated according to a proper continualization technique [55]. In contrast, the Lagrangian
model accurately captures the positive experimental ratio E12 E22 (yellow circles in Figure 8c),
which measures the coupling between the normal strain E22 and the angular strain E12 .

3. BI-TETRACHIRAL SAMPLE
3.1 Sample preparation, Experimental set-up, Data acquisition and processing
The bi-layered polymeric sample of the bi-tetrachiral geometry has been obtained by simply
superimposing two tetrachiral samples (front and back layers) with opposite chirality. The
geometric dimensions of the printed sample are reported in Table 2.
As for the tetrachiral samples, the bi-tetrachiral samples made of an array of 6-by-6 cells
have been experimentally tested under uni-axial tension according to the force-control scheme,
within stable environmental conditions. Known increments of force have been applied to the
sample, starting from the initial constrained configuration (Figure 10a), taken as reference. In
the reference configuration, the sample lies in the vertical plane under self-weight load and is
supported by the alignment-preserving constraining system. The supporting system is designed

18
to constrain the relative displacements (but not the relative rotation) of all the 6 ring pairs (one
ring in the front layer, the other in the back layer) at the top side. The steel truss system acting
at the bottom side of the bi-tetrachiral sample applies the controlled force and simultaneously
constrains the relative displacements (but not the relative rotation) of all the 6 ring pairs (one
ring in the front layer, the other in the back layer) at the bottom side. The total vertical action
is split into equal forces applied at all the ring pairs located in the last (bottom) row of the
sample (Figure 10b). Apart from the relative constraints at the top and bottom sides, each layer
is free to independently develop both lateral expansion/contraction and the rolling-up
mechanism. In order to assess the global elastic properties of the bi-tetrachiral sample, the 4-
by-6 inner cluster of internal unconstrained and unloaded cells has been considered in the
following.
The uni-axial tension test has been run by applying five increments of the quasi-static force
F2 , corresponding to equivalent global stress  22  F2 A2 , where A2  12 Hd is the cross-
section area of the ideal solid with dimensions 6H (width), 6H (height) and 2d (depth). The
force increments (steps 1-5) are the same already reported in Table 2, starting from the initial
configuration (step 0). Attention has been paid to preserve the vertical planarity of the deformed
configurations, minimizing as much as possible the undesired effects of friction in the
constraints and not overcoming the limit of linear reversible deformation in the material.
As for the tetrachiral samples, the non-contact data acquisition has been performed through
a digital camera. Specifically, one or more two-dimensional images (with dimensions 5184-by-
3456 pixels) of the bi-tetrachiral specimen have been acquired for each load step, taking care
of minimizing distortion by properly calibrating the focal axis and focal plane of the camera
according to the specimen plane and geometry. After uploading the digital photographs in the
Matlab environment, the Image Processing Toolbox has been employed to convert the pixel
coordinates of the digital photograph into a real coordinate system. The real positions x of the
ring centers of the layers have been again assessed by recognizing the circular shapes by virtue
of the Hough transform for each load step. Therefore, the total displacement vector v of the
ring centers have been calculated as position differences with respect to the initial load step.
Finally, the global elastic properties (Young modulus E and Poisson ratio  ) are assessed
following the procedure described in Section 1.3.2.

3.2 Numerical simulations and analytical predictions

3.2.1 Solid finite element model
In analogy with the tetrachiral samples, the bi-tetrachiral samples can be modelled according to
a high-dimensional formulation in the framework of linear solid mechanics. The two layers
have been first reproduced with high geometrical fidelity in the framework of the CAD tool
[62] (Figure 11a illustrates the six-by-six cell sample), and successively discretized and

19
(a) (b)

Figure 11. Three-dimensional solid model of the bi-tetrachiral sample: (a) perspective geometric view
of the six-by-six cell sample; (b) finite element mesh and detail of the periodic cell.

analyzed using Abaqus [63]. The same mesh size and the element type of the tetrachiral solid
model have been adopted for the two layers of the bi-tetrachiral samples (for instance, the six-
by-six cell sample has been meshed in 129834 C3D8R elements and the mesh detail of the
periodic cell are illustrated in Figure 11b).
The solid model has been specified by assuming the same isotropic homogeneous material,
the same rigid body assumptions and the same external boundary conditions of the tetrachiral
model. In addition, inter-layer constraints have been defined to impose the cell-to-cell identity
between: (i) the pairs of horizontal displacements of the ring centers in the first (top) row of the
two layers, (ii) the pairs of horizontal displacements and the pairs of vertical displacements of
the ring centers in the last (bottom) row of the two layers. Equal vertical forces have been
applied to each centroid of all the rigid bodies of the last (bottom) cell row in the front and back
layer. From the finite element solution, the 48-by-one displacement subvector v that collects
the horizontal and vertical displacement components of the ring centers in the inner cluster of
4-by-6 cells has been reconstructed and then employed to identify the equivalent elastic
parameters.

3.2.2 Beam lattice model

In analogy with the tetrachiral samples, the bi-layered periodic cell of the bi-tetrachiral samples
can again be modelled according to a low-dimensional lagrangian formulation (Figure 12a).
The independent geometric properties of the periodic square cell are the side length H and the
ring radius R , while the chirality angles of the tangent ligaments are    arctan  2 R H  ,
where the positive and the negative signs correspond to the front layer (identified by the
superscript (1) in the following) and to the back one (superscript (2)), respectively (Figure 12b).

20
(a) (b) (c) q5(1)
First
Row q5(2)
Front  Back
layer layer q(2)
2
q(1)
2
q(1)
4
N rows

H
  q1(2) q(2)
4
q1(1)
Last 
Row q3(1)

M columns H q3(2)

Figure 12. Bi-Tetrachiral sample: (a) repetitive planar pattern, (b) periodic cell, (c) beam lattice model.

Based on the simplifying mechanical assumptions already introduced for the lagrangian
model of the tetrachiral samples, identical structural properties (stiff rings, flexible ligaments
with Young modulus E , axial and flexural rigidities EA and EJ ) are considered for the two
layers of the bi-tetrachiral samples. In the reference planar configuration, the two internal nodes
(one for each layer) and the eight external nodes (four for each layer) share the same position.
The deformed configuration of the periodic cell is fully described by three planar degrees-
of-freedom for each configurational node (Figure 12c). The degrees-of-freedom of the internal
node and the i-th external node (with i  1,..., 4 ) of the k -th layer are collected column-wise in

the two vectors q1  u1 , v1 ,1
(k ) (k ) (k ) (k )
 (k )

and eight vectors qi  ui , vi ,i , respectively. All
(k ) (k ) (k )

the nodal degrees-of-freedom are collected in the two 15-by-1 cellular displacement vectors

q(1)  q1(1) , q(1) (1) (1)

2 , q3 , q4 , q5
(1) (2) (2)
 (2) (2) (2) (2)

and q  q1 , q2 , q3 , q4 , q5 . The 15-by-15 stiffness matrix
K (1) of the generic cell in the front layer is the same already presented for the tetrachiral
periodic cell in Appendix A, while the 15-by-15 stiffness matrix K ( 2 ) of the generic cell in the
back layer is obtained straightforwardly by changing the  -sign.
Considering a bidimensional array of N  M cells (with N rows and M columns), the
global 3P -by-one displacement vectors q (1) g and q (2)
g can be defined for two layers of the bi-
tetrachiral samples. Introducing the cellular-to-global change-of-coordinates q (jk )  Λ j q (gk ) for
the j -th cell (with j  1,..., NM ), the 3P -by- 3P global stiffness matrices read
NM NM

g  Λ jK Λ j,
K (1) g  Λ jK
T (1)
K (2) T (2)
Λj (13)
j 1 j 1

where the 15-by- 3P boolean allocation matrices Λ j hold for both the layers. Joining the layer
displacement vectors into the 6P -by-1global displacement vector q g  q g , q g , the 6P -
(1) (2)
 
by- 6P stiffness matrix of the bi-tetrachiral sample reads
 K (1) O 
Kg   g (2) 
(14)
 O K g 
21
Consequently, apart the doubled dimension, the algebraic equations governing the static
equilibrium of the unconstrained lagrangian model are formally identical to the equations of the
tetrachiral samples
K gqg  fg (15)

where fg is the 6P -by-one vector of the forces acting in the components of the vector q g .
In order to describe the boundary conditions at the top side of the bi-tetrachiral sample,
proper external constraints of the central nodes of the first (top) cell row must be imposed.
Defining the 2  M  1  -by-one displacement vector qc that collects all the vertical
displacements (plus a pair of horizontal displacements, one for each layer, to avoid rigid motion
solutions) of the central nodes in the 2M cells of the first (top) row, the following partition can
be introduced
K uu K uc   qu   fu 
K K   q    f  (16)
 cu cc   c   c 

where the 2  3 P  M  1 -by-one vector qu collects all the externally unconstrained degrees-
of-freedom. Internal constrains are also necessary to describe the inter-layer coupling at the top
and bottom boundaries. To this purpose, the unconstrained displacement vector can be
decomposed as qu   q m , q s  , where the 6  P  M  -by-one master displacement vector qm and
the 2  2 M  1 -by-one slave displacement vector qs are distinguished. Therefore, the internal
constraints can be imposed on the slave vector through the relation qs  Vqm , where the
2  2 M  1 -by- 6  P  M  boolean constraint matrix V imposes the cell-to-cell identity
between: (i) the pairs of horizontal displacements of the central nodes in the first (top) row of
the two layers, (ii) the pairs of horizontal displacements and the pairs of vertical displacements
of the central nodes in the last (bottom) row of the two layers. Accordingly, the force vector
can be decomposed f u   f m , f s  . Consequently, the equation (16) can be partitioned in the
convenient form
K mm K ms K mc   qm   fm 
K K K  q    f 
 sm ss sc   s   s (17)
 K cm K cs K cc   qc   fc 
 

and, imposing the constraint condition qc  0 , the equation (17) can be inverted to obtain the
static equilibrium solution

q m   K mm  K ms V  f m
1

f s   K sm  K ss V  K mm  K ms V  f m
1
(18)
f c   K cm  K cs V  K mm  K ms V  f m
1

22
 Figure 13. Finite element simulation of the bi-tetrachiral sample: (a) color map of the horizontal
displacement component; (b),(c) comparison between the undeformed (gray lines) and deformed
configuration (yellow and green solids).

where fm has non-null equal components corresponding to the only vertical forces of the central
nodes in the last (bottom) row of the two layers. Again, in order to extract from the displacement
solution qm the pseudo-experimental results necessary for the identification of the equivalent
elastic parameters, the 48-by-one displacement subvector v that collects the horizontal and
vertical displacement of the central nodes in the inner cluster of 4-by-6 cells is considered.

3.3 Comparison of experiments and simulations

The experimental results are compared qualitatively and quantitatively with the simulations
obtained with the lagrangian and the finite element solid model of the bi-tetrachiral sample.
First, from a qualitative point of view, the two layers exhibit a different but mutually and doubly
symmetric behaviour in all the experimental deformed configurations under the symmetric
scheme of external forces. Specifically, the response of the two-layer sample is symmetric with
respect to the vertical and horizontal in-plane axes crossing the sample barycenter. The double
symmetry of the experimental response is confirmed by the static simulations. This peculiar
behaviour can be essentially attributed to the opposite chiralities of the two layers composing
the cellular topology. The opposite chirality angles determine the concurrent development of
two independent mechanisms of local deformation, one for each layer, activated by the
opposite-sign rotations of the rings around their centers. Globally, the vertical stretching of the
sample – collinear to the external force direction – is not accompanied with any appreciable
angular strain. This characteristic behaviour is observable in the experimental deformed confi-
gurations, as well as in the numerical simulations obtained from the solid model (Figure 13).
23
(a) (b) (c)

Figure 14. Comparison between experimental results (red circles) and numerical results of the solid model
(green circles) and beam lattice model (yellow circles): (a) equivalent global stress  2 2 versus normal strain
 2 2 , (b) normal strain 11 versus normal strain  2 2 , (c) angular strain 12 versus normal strain  2 2

The horizontal component of the normalized displacement field U Umax is shown in Figure
13a, while the deformed configurations of the front layer (yellow) and back layer (green) are
portrayed in Figure 13b,c. Second, from a quantitative point of view, the normal and angular
strains identified from the experimental data are compared with the corresponding strains
identified from the numerical results obtained with the lagrangian and the finite element solid
models (Figure 14), respectively. As for the tetrachiral sample, it can be noted that the
experimental response (red circles) shows a quasi-perfectly linear behaviour under increasing
values of the external forces.
With focus on the solid model, which employs the updated Young modulus of the ABS
material (1540 MPa), the ratio  22 E22 obtainable from the numerical results of the updated
solid model (green circles in Figure 14a) shows a fine agreement – apart from a minor
underestimation – with the corresponding experimental ratio (red circles in Figure 14a).
Consistently with this remark, the global Young modulus ( E  2.11 MPa) identified from the
numerical results (green circles in Figure 15a) is slightly lower than the minimum and the mean
value ( E  2.26 MPa) of the global Young modules identified from the experimental results
(circles reds in Figure 15a). In synthesis, it can be highlighted how the global Young modulus
of the bi-tetrachiral sample – in virtue of the two layer collaboration – is about twice that of the
tetrachiral sample made of the same material.
Adopting the updated solid model, a very satisfactory agreement is again achieved between
the ratio E11 E22 obtained from the numerical results (green circles in Figure 14b) and the
corresponding experimental ratio (red circles in Figure 14b). However, the positive values
systematically attained by this ratio for increasing external forces demonstrate a remarkably
auxetic behaviour of the bi-tetrachiral sample in the direction orthogonal to the external forces.
This key result fundamentally differentiates the global behaviour of the bi-tetrachiral material
from that of the tetrachiral material individually characterizing the two component layers. From
24
 (a) (b)

Figure 15. Comparison between experimental results (red circles) and numerical results of the solid
model (green circles) and beam lattice model (yellow circles). Global elastic properties versus external
load F 2 2 : (a) Young modulus E , (b) Poisson ratio  .

a mechanical perspective, this drastic change in the static response can be fully attributed to the
inter-layer rigid constraint at the boundaries (top and bottom cell rows). The auxetic behaviour
is effectively synthesized by the negative values systematically identified for the global Poisson
ratio  (Figure 15b). In particular, the load-independent value (   0.71 ) identified from the
numerical results (green circles in Figure 15b) effectively matches – apart from a small
underestimation – the mean value (   0.64 ) identified from the experimental results (red
circles in Figure 15b). Finally, the first qualitative remark concerned with the absence of an
appreciable angular strain is quantitatively confirmed by the identification of the angular strain
E12 , which assumes experimental and numerical values in mutual agreement but significantly
lower (by two orders of magnitude) than those of the normal strain E22 (green and red circles
in Figure 14c).
With focus on the lagrangian model, a satisfying agreement is found in the simulation
(yellow circles in Figure 14a) of the experimental ratio  22 E22 . The identification of the
global Young modulus from the lagrangian simulations (load-independent values marked by
yellow circles in Figure 15a) returns a slight underestimation ( E  1.98 MPa) of the
corresponding values identified from the experiments (red circles) and the finite element solid
model (green circles). Similarly to the tetrachiral sample, this difference can be attributed to the
simplifying assumptions of the lagrangian formulation [55]. Despite the simplifying
assumptions, the lagrangian formulation allows the identification of positive values for the ratio
E11 E22 , consistently with the experimental results. Nonetheless, it can be noted that the
corresponding load-independent identification of the global Poisson ratio (   0.95 marked

25
 (a) (b)

Figure 16. Comparison between experimental results (red circle) and numerical results of the solid model
(green circles) and beam lattice model (yellow circles). Global elastic properties versus cell number of the
square sample (with N  M ): (a) Young modulus E , (b) Poisson ratio  .

by yellow circles in Figure 15b) systematically overestimates the actual auxeticity of the bi-
tetrachiral sample. Finally, the lagrangian model confirms a certain accuracy in fitting the
experimental ratio E12 E22 (yellow circles in Figure 14c).
Once the actual reliability of the mechanical formulations in simulating the static response
of the bi-tetrachiral material has been experimentally verified, the solid and the lagrangian
models have been used to simulate the behaviour of square samples characterized by increasing
size (up to N  M 18 ). Given the key role played by the inter-layer boundary constraints on the
auxetic response of the bi-tetrachiral material, the purpose of these simulations is targeted to
verifying the influence of the boundary-to-boundary distance on the global elastic properties.
Regardless of the sample size, the identification of the global Young modulus and global
Poisson ratio has been based on the displacement subvector v related to the inner cluster of 4-
by-6 cells, for the sake of consistency. The values of the global Young modulus identified from
the solid model (green circles) and from the lagrangian model (yellow circles) exhibit limited
variability for increasing sizes of the samples (Figure 16a). Specifically, E -values ranging
from 2.08 MPa to 2.14 MPa have been obtained for the solid model (respectively for N  8 and
N  14 ), while E -values ranging from 1.94 MPa and 1.98 MPa have been obtained for the
lagrangian model (respectively for N  8 and N  6 ). Differently, the values of the global
Poisson ratio identified from the solid model (green circles) and from the lagrangian model
(yellow circles) exhibit a monotonically convergent behaviour for increasing sizes of the
samples. For the largest size ( N  18 ) the global Poisson ratio attains the minimum values
  0.99 for the solid model and   1.10 for the lagrangian model (Figure 16b). This

26
persistent result tends to confirm that the auxeticity is a property of the bi-tetrachiral material
that actually leverages the inter-layer boundary constraints, but does not vanish for large
boundary-to-boundary distances.

CONCLUSIONS
The global elastic properties of architectured honeycomb materials characterized by a
tetrachiral cellular microstructure, made of circular stiff rings tangentially connected by flexible
ligaments, have been investigated. The theoretical predictions based on low-dimension beam
lattice models and finite element solid formulations have been compared with the experimental
data obtained from quasi-static laboratory tests performed on different planar samples realized
with the 3D printing technology. As major experimental evidence, the tetrachiral samples have
shown a remarkable coupling between the normal and angular strains under increasing tension
loads, due to the local ring rotations activated by the cellular chirality (rolling up mechanism).
For the particular test direction, the global Young modulus and the global (positive) Poisson
ratio of the tetrachiral material samples, experimentally identified by solving an overdetermined
inverse problem, are accurately predicted by the numerical results of the solid computational
model. On the contrary, some simplifying mechanical assumptions tend to limit the actual
descriptive possibilities of the beam lattice model, particularly if the ring deformability is
comparable with the ligament flexibility.
Inspired by the encouraging findings concerning the tetrachiral material, an original bi-
layered topology of microstructured tetrachiral materials (bi-tetrachiral materials) has been
theoretically conceived and mechanically modelled. The bi-tetrachiral topology exploits the
virtuous mutual collaboration of two tetrachiral layers with opposite chirality in order to prevent
the development of angular strains under tensile loads. As major technological advantage, the
new material topology does not require a different bi-layered multi-material 3D printing
process. The angular strain elimination is kinematically based on the independent and opposite-
sign rolling up mechanisms of the two component layers, reciprocally constrained at the sample
boundaries. Specifically, the boundary constraints are properly designed to impose identical
layer-to-layer displacements to the rings, which remain free to rotate independently.
As main macroscopic consequence of its microstructural layout, the static response of the
bi-tetrachiral material – composed of collaborating non-auxetic layers – exhibits a remarkably
strong auxeticity that can be quantified by negative global Poisson ratio close to -0.7 for the
tested samples. Moreover, the bi-tetrachiral material is found to outperform the tetrachiral
material also in terms of global Young modulus, which turns out to be nearly doubled under the
same testing conditions. These experimental results have been closely confirmed by the
qualitative and quantitative simulations obtainable with beam lattice models and finite element
solid formulations. In this case, the simplifying mechanical assumptions affecting the beam

27
lattice model determine only a minor underestimation of the global Young modulus and the
global Poisson ratio. Finally, parametric analyses have been carried out to evaluate numerically
the effects of the bi-tetrachiral sample size on the global elastic properties. The result trend has
demonstrated that a minimum Poisson ratio close to -1 can actually be reached by increasing
the distance between the sample boundaries providing the interlayer constraint. This systematic
results tend to confirm that the strong auxetic behaviour is substantially independent of the
sample size, that is, represents a characteristic property of the bi-tetrachiral material.

ACKNOWLEDGEMENTS
The authors ML, FV and LG acknowledge financial support of the (MURST) Italian
Department for University and Scientific and Technological Research in the framework of the
research MIUR Prin15 project 2015LYYXA8 entitled Multi-scale mechanical models for the
design and optimization of micro-structured smart materials and metamaterials, coordinated
by prof. A. Corigliano. The author AB acknowledges financial support by National Group of
Mathematical Physics (GNFM-INdAM). The authors FA and SM acknowledge the partial
support by the strategic project of the University of Pavia entitled Virtual Modeling and
Additive Manufacturing for Advanced Materials (3D@UniPV). Finally, all the authors
acknowledge Dr. Gianluca Aliamo for technical support in printing all the samples, and Dr.
Giuseppe Riotto for the assistance in designing and performing the experimental tests.

COMPLIANCE WITH ETHICAL STANDARDS

The authors declare that they have no conflict of interest

REFERENCES
[1] Fleck, N.A., Deshpande, V.S., Ashby, M.F. (2010). Micro-architectured materials: past,
present and future. Proceedings of the Royal Society A: Mathematical, Physical and
Engineering Sciences, 466(2121), 2495-2516.
[2] Schaedler, T.A., Carter, W.B. (2016). Architected cellular materials. Annual Review of
Materials Research, 46, 187-210.
[3] Meza, L.R., Zelhofer, A.J., Clarke, N., Mateos, A.J., Kochmann, D.M., Greer, J.R.
(2015). Resilient 3D hierarchical architected metamaterials. Proceedings of the National
Academy of Sciences, 112(37), 11502-11507.
[4] Chen, Y.J., Scarpa, F., Liu, Y.J., Leng, J.S. (2013). Elasticity of anti-tetrachiral
anisotropic lattices. International Journal of Solids and Structures, 50(6), 996-1004.

28
[5] Saxena, K.K., Das, R., Calius, E.P. (2016). Three decades of auxetics research − materials
with negative Poisson's ratio: a review. Advanced Engineering Materials, 18(11), 1847-
1870.
[6] Ha, C.S., Plesha, M.E., Lakes, R.S. (2016). Chiral three‐dimensional isotropic lattices
with negative Poisson's ratio. Physica status solidi (b), 253(7), 1243-1251.
[7] Lakes, R.S. (2017). Negative-Poisson's-Ratio Materials: Auxetic Solids. Annual Review
of Materials Research, 47, 63-81.
[8] Ren, X., Das, R., Tran, P., Ngo, T.D., Xie, Y.M. (2018). Auxetic metamaterials and
structures: A review. Smart materials and structures, 27(2), 023001.
[9] Duncan, O., Shepherd, T., Moroney, C., Foster, L., Venkatraman, P., Winwood, K., Allen
T., Alderson, A. (2018). Review of auxetic materials for sports applications: Expanding
options in comfort and protection. Applied Sciences, 8(6), 941.
[10] Spadoni, A., Ruzzene, M., Gonella, S., Scarpa, F. (2009). Phononic properties of
hexagonal chiral lattices. Wave motion, 46(7), 435-450.
[11] Tee, K.F., Spadoni, A., Scarpa, F., Ruzzene, M. (2010). Wave propagation in auxetic
tetrachiral honeycombs. Journal of Vibration and Acoustics, 132(3), 031007.
[12] Lepidi, M., Bacigalupo, A. (2018). Parametric design of the band structure for lattice
materials. Meccanica, 53(3), 613-628.
[13] Vadalà, F., Bacigalupo, A., Lepidi, M., Gambarotta, L. (2018). Bloch wave filtering in
tetrachiral materials via mechanical tuning. Composite Structures, 201, 340-351.
[14] Bacigalupo, A., Lepidi, M., Gnecco, G., Vadalà, F., Gambarotta, L. (2019). Optimal
design of the band structure for beam lattice metamaterials. Frontiers in Materials, 6, 2.
[15] Bettini, P., Airoldi, A., Sala, G., Di Landro, L., Ruzzene, M., Spadoni, A. (2010).
Composite chiral structures for morphing airfoils: Numerical analyses and development
of a manufacturing process. Composites Part B: Engineering, 41(2), 133-147.
[16] Liu, X.N., Hu, G.K., Sun, C.T., Huang, G.L. (2011). Wave propagation characterization
and design of two-dimensional elastic chiral metacomposite. Journal of Sound and
Vibration, 330(11), 2536-2553.
[17] Ranjbar, M., Boldrin, L., Scarpa, F., Neild, S., Patsias, S. (2016). Vibroacoustic
optimization of anti-tetrachiral and auxetic hexagonal sandwich panels with gradient
geometry. Smart Materials and Structures, 25(5), 054012.
[18] Bacigalupo, A., Gambarotta, L. (2017). Wave propagation in non-centrosymmetric beam-
lattices with lumped masses: Discrete and micropolar modeling. International Journal of
Solids and Structures, 118, 128-145.
[19] Tallarico, D., Movchan, N.V., Movchan, A.B., Colquitt, D.J. (2017). Tilted resonators in
a triangular elastic lattice: chirality, Bloch waves and negative refraction. Journal of the
Mechanics and Physics of Solids, 103, 236-256.

29
[20] Bacigalupo, A., Lepidi, M. (2018). Acoustic wave polarization and energy flow in
periodic beam lattice materials. International Journal of Solids and Structures, 147, 183-
203.
[21] Lakes, R. (1991). Deformation mechanisms in negative Poisson's ratio materials:
structural aspects. Journal of materials science, 26(9), 2287-2292.
[22] Prall, D., Lakes, R.S. (1997). Properties of a chiral honeycomb with a Poisson's ratio of -
1. International Journal of Mechanical Sciences, 39(3), 305-314.
[23] Alderson, A., Alderson, K.L., Attard, D., Evans, K.E., Gatt, R., Grima, J.N., Miller, W.,
Ravirala, N., Smith, C.W., Zied, K. (2010). Elastic constants of 3-, 4-and 6-connected
chiral and anti-chiral honeycombs subject to uniaxial in-plane loading. Composites
Science and Technology 70(7), 1042-1048.
[24] Lorato, A., Innocenti, P., Scarpa, F., Alderson, A., Alderson, K.L., Zied, K.M., Ravirala
N., Miller W., Smith C.W., Evans, K.E. (2010). The transverse elastic properties of chiral
honeycombs. Composites Science and Technology 70(7), 1057-1063.
[25] Alderson, A., Alderson, K.L., Chirima, G., Ravirala, N., Zied, K.M. (2010). The in-plane
linear elastic constants and out-of-plane bending of 3-coordinated ligament and cylinder-
ligament honeycombs. Composites Science and Technology 70(7), 1034-1041.
[26] Miller, W., Smith, C.W., Scarpa, F., Evans, K.E. (2010). Flatwise buckling optimization
of hexachiral and tetrachiral honeycombs. Composites Science and Technology 70(7),
1049-1056.
[27] Abramovitch, H., Burgard, M., Edery-Azulay, L., Evans, K.E., Hoffmeister, M., Miller,
W., Scarpa F., Smith C.W., Tee, K.F. (2010). Smart tetrachiral and hexachiral
honeycomb: Sensing and impact detection. Composites Science and Technology 70(7),
1072-1079.
[28] Li, H., Ma, Y., Wen, W., Wu, W., Lei, H., Fang, D. (2017). In plane mechanical properties
of tetrachiral and antitetrachiral hybrid metastructures. Journal of Applied Mechanics,
84(8), 081006.
[29] Ma, C., Lei, H., Hua, J., Bai, Y., Liang, J., Fang, D. (2018). Experimental and simulation
investigation of the reversible bi-directional twisting response of tetra-chiral cylindrical
shells. Composite Structures, 203, 142-152.
[30] Wu, W., Geng, L., Niu, Y., Qi, D., Cui, X., Fang, D. (2018). Compression twist
deformation of novel tetrachiral architected cylindrical tube inspired by towel gourd
tendrils. Extreme Mechanics Letters, 20, 104-111.
[31] Li, M., Lu, X., Zhu, X., Su, X., Wu, T. (2018). Research on in‐plane quasi‐static
mechanical properties of gradient tetra‐chiral hyper‐structures. Advanced Engineering
Materials, 1801038, doi: 10.1002/adem.201801038.

30
[32] Jiang, Y., Li, Y. (2018). Novel 3d‐printed hybrid auxetic mechanical metamaterial with
chirality‐induced sequential cell opening mechanisms. Advanced Engineering Materials
20(2), 1700744.
[33] Fu, M., Liu, F., Hu, L. (2018). A novel category of 3D chiral material with negative
Poisson's ratio. Composites Science and Technology, 160, 111-118.
[34] Wu, W., Qi, D., Liao, H., Qian, G., Geng, L., Niu, Y., Liang, J. (2018). Deformation
mechanism of innovative 3D chiral metamaterials. Scientific reports, 8(1), 12575.
[35] Guo, N., Leu, M.C. (2013). Additive manufacturing: technology, applications and
research needs. Frontiers of Mechanical Engineering, 8(3), 215-243.
[36] Gao, W., Zhang, Y., Ramanujan, D., Ramani, K., Chen, Y., Williams, C.B., Wang C.C.L.,
Shin Y.C., Zhang S., Zavattieri, P.D. (2015). The status, challenges, and future of additive
manufacturing in engineering. Computer-Aided Design, 69, 65-89.
[37] Jiang, Y., Li, Y. (2017). 3D printed chiral cellular solids with amplified auxetic effects
due to elevated internal rotation. Advanced Engineering Materials, 19(2), 1600609.
[38] Chen, D., Zheng, X. (2018). Multi-material additive manufacturing of metamaterials with
giant, tailorable negative Poisson’s ratios. Scientific reports, 8, 9139.
[39] Clausen, A., Wang, F., Jensen, J.S., Sigmund, O., Lewis, J.A. (2015). Topology
optimized architectures with programmable Poisson's ratio over large deformations.
Advanced Materials, 27(37), 5523-5527.
[40] Schaedler, T.A., Carter, W. B. (2016). Architected cellular materials. Annual Review of
Materials Research, 46, 187-210.
[41] Muth, J.T., Dixon, P.G., Woish, L., Gibson, L.J., Lewis, J.A. (2017). Architected cellular
ceramics with tailored stiffness via direct foam writing. Proceedings of the National
Academy of Sciences, 114(8), 1832-1837.
[42] Dong, G., Tang, Y., Zhao, Y.F. (2017). A survey of modeling of lattice structures
fabricated by additive manufacturing. Journal of Mechanical Design, 139(10), 100906.
[43] Maskery, I., Sturm, L., Aremu, A.O., Panesar, A., Williams, C.B., Tuck, C.J., Wildman
R.D., Ashcroft I.A., Hague, R.J.M. (2018). Insights into the mechanical properties of
several triply periodic minimal surface lattice structures made by polymer additive
manufacturing. Polymer, 152, 62-71.
[44] Vyatskikh, A., Delalande, S., Kudo, A., Zhang, X., Portela, C.M., Greer, J.R. (2018).
Additive manufacturing of 3D nano-architected metals. Nature communications, 9(1),
593.
[45] Vayre, B., Vignat, F., Villeneuve, F. (2012). Metallic additive manufacturing: state-of-
the-art review and prospects. Mechanics & Industry, 13(2), 89-96.

31
[46] Sha, Y., Jiani, L., Haoyu, C., Ritchie, R.O., Jun, X. (2018). Design and strengthening
mechanisms in hierarchical architected materials processed using additive manufacturing.
International Journal of Mechanical Sciences, 149, 150-163.
[47] Melde, K., Mark, A.G., Qiu, T., Fischer, P. (2016). Holograms for acoustics. Nature,
537(7621), 518.
[48] Auricchio, F., Marconi, S. (2016). 3D printing: clinical applications in orthopaedics and
traumatology. EFORT open reviews, 1(5), 121-127.
[49] Capel, A.J., Rimington, R.P., Lewis, M.P., Christie, S.D. (2018). 3D printing for
chemical, pharmaceutical and biological applications. Nature Reviews Chemistry, 2, 422-
436.
[50] Macdonald, E., Salas, R., Espalin, D., Perez, M., Aguilera, E., Muse, D., Wicker, R.B.
(2014). 3D printing for the rapid prototyping of structural electronics. IEEE access, 2,
234-242.
[51] Massoni, E., Silvestri, L., Bozzi, M., Perregrini, L., Alaimo, G., Marconi, S., Auricchio,
F. (2016). Characterization of 3D-printed dielectric substrates with different infill for
microwave applications. In 2016 IEEE MTT-S Int. Microwave Workshop Series on
Advanced Materials and Processes for RF and THz Applications (IMWS-AMP).
[52] Cui, H., Hensleigh, R., Yao, D., Maurya, D., Kumar, P., Kang, M.G., Priya  S., Zheng,
X.R. (2019). Three-dimensional printing of piezoelectric materials with designed
anisotropy and directional response. Nature materials, 18(3), 234-241.
[53] Huang, S.H., Liu, P., Mokasdar, A., Hou, L. (2013). Additive manufacturing and its
societal impact: a literature review. The International Journal of Advanced Manufacturing
Technology, 67(5-8), 1191-1203.
[54] Auricchio, F., Greco, A., Alaimo, G., Giacometti, V., Marconi, S., and Mauri, V. (2017).
3D printing technology for buildings accessibility: the tactile map for MTE museum in
Pavia, Journal of Civil Engineering and Architecture, 11, 736-747.
[55] Bacigalupo, A., Gambarotta, L. (2014). Homogenization of periodic hexa-and tetrachiral
cellular solids. Composite Structures, 116, 461-476.
[56] Chen, Y., Liu, X.N., Hu, G.K., Sun, Q.P., Zheng, Q.S. (2014). Micropolar continuum
modelling of bi-dimensional tetrachiral lattices. Proceedings of the Royal Society A:
Mathematical, Physical and Engineering Sciences, 470(2165), 20130734.
[57] Cabras, L., Brun, M. (2014). Auxetic two-dimensional lattices with Poisson's ratio
arbitrarily close to -1. Proceedings of the Royal Society of London Series A, 470,
20140538-20140538.
[58] Coble, S. (2003) Materials data book. Cambridge University Engineering Department,
Cambridge, UK.

32
[59] Gil, S., Reisin, H.D., Rodríguez, E.E. (2006). Using a digital camera as a measuring
device. American journal of physics, 74(9), 768-775.
[60] MATLAB Full Suite (2019), The Math Works Inc., Natick, USA.
[61] Yuen, H.K., Princen, J., Illingworth, J., Kittler, J. (1990). Comparative study of Hough
transform methods for circle finding. Image and vision computing, 8(1), 71-77.
[62] Autodesk Inventor (2017), Autodesk, San Rafael, USA.
[63] Abaqus (2017), Abaqus Theory Manual, SIMULIA - Dassault Systèmes, Providence,
USA.

APPENDICE A
A.1 Structural matrices
The 15-by-15 stiffness matrix K of the periodic cell defined in Section 2.2 for the tetrachiral
sample can be expressed in the form
 K 11 K12 K 13 K14 K15 
 K 22 O O O 
K   ... K 33 O O  (19)
 K 44 O 
Sym ... K 55 

where the three-by-three non-null submatrices are

C 0 0 
2 1
K 11  3 0 C1 0  (20)
L  2

 0 0 2C 2 L 

 EAL4  6C3 R2 2 2C4 LR1 C5 L2 R 2 

 
K12  2 22 L5  2 2C4 LR1 4C6 R2  6EJL2  2C7 L3 2  (21)
 C5 L2 R 2 2C7 L3 2 C8 L4 22 


4C6 R2  6EJL2 2 2C4 LR1 2C7 L3 2 

 
K13  2 22 L5  2 2C4 LR1  EAL4  6C3 R2 C5 L2 R 2  (22)
  2C7 L3 2 C5 L2 R 2 C8 L4 22 


 EAL4  6C3 R2 2 2C4 LR1 C5 L2 R2 

 
K14  222 L5  2 2C4 LR1 4C6 R2  6EJL2 2C7 L3 2  (23)
 C5 L2 R2  2C7 L32 C8 L422 


33
 4C6 R2  6EJL2 2 2C4 LR1  2C7 L3 2 
 
K15  2 22 L5  2 2C4 LR1  EAL4  6C3 R2 C5 L2 R 2  (24)
 2C7 L3 2 C5 L2 R 2 C8 L4 22 


 EAL4  6C3 R2 2 2C4 LR1 C5 L2 R 2 

 
K 22  2 22 L5  2 2C4 LR1 4C6 R2  6EJL2 2C7 L3 2  (25)
 C5 L2 R 2 2C7 L3 2 C2 L4 22 


4C6 R2  6EJL2 2 2C4 LR1  2C7 L3 2 

 
K 33  2 22 L5  2 2C4 LR1 EAL4  6C3 R2 C5 L2 R 2  (26)
  2C7 L3 2 C5 L2 R 2 C2 L4 22 


 EAL4  6C3 R2 2 2C4 LR1 C5 L2 R2 

 
K 44  2 22 L5  2 2C4 LR1 4C6 R2  6EJL2  2C7 L3 2  (27)
 C5 L2 R2  2C7 L3 2 C2 L422 


4C6 R2  6EJL2 2 2C4 LR1 2C7 L3 2 

 
K 55  2 22 L5  2 2C4 LR1 EAL4  6C3 R2 C5 L2 R 2  (28)
 2C7 L3 2 C5 L2 R 2 C2 L4 22 


6R2 2R2
with the auxiliary coefficients 1  1  and  2  1  and the auxiliary parameters
L2 L2
C1  EAL2  6 EJ
C2  EAR 2  4 EJ
C3  EAL2  8EJ
C4  EAL2  6 EJ
(29)
C5  EAL21  12EJ
C6  2 EAL2  9 EJ
C7  2EAR 2 +3EJ1
C8  EAR 2  2 EJ

34